J Manipulative Physiol Ther 2004 (Mar); 27 (3): 160–169 ~ FULL TEXT
Mitchell Haas, DC, Bruce Goldberg, MD, Mikel Aickin, PhD, Bonnie Gangera, Michael Attwood
Mitchell Haas, DC,
Center for Outcome Studies,
Western States Chiropractic College,
2900 NE 132nd Ave,
Portland, OR 97230;
Objective: This study reports pain and disability outcomes up to 4 years for chiropractic and medical patients with low back pain (LBP) and assesses the influence of doctor type and pain duration on clinical outcomes.
Design: Prospective, longitudinal, nonrandomized, practice-based, observational study.
Setting: Fifty-one chiropractic and 14 general practice community clinics.
Subjects: A total of 2870 acute and chronic ambulatory patients with LBP of mechanical origin.
Methods: Sixty chiropractic (DC) and 111 general practice (MD) physicians participated. Primary outcomes were pain, using a 100-point visual analogue scale (VAS), and functional disability, using the Revised Oswestry Disability Questionnaire. These were measured at baseline and 8 time points. Regression analysis compared acute and chronic DC and MD patients after correcting for baseline differences in the 4 cohorts.
Results: Most improvement was seen by 3 months and sustained for 1 year; exacerbation was seen thereafter. Acute patients demonstrated greater relief at all time points. A clinically important advantage for chiropractic patients was seen in chronic patients in the short-term (>10 VAS points), and both acute and chronic chiropractic patients experienced somewhat greater relief up to 1 year (P< .000). The advantage for DC care was prominent for chronic patients with leg pain below the knee (P< .001). More than 50% of chronic patients had over 50 days of pain in the third year.
Conclusion: Study findings were consistent with systematic reviews of the efficacy of spinal manipulation for pain and disability in acute and chronic LBP. Patient choice and interdisciplinary referral should be prime considerations by physicians, policymakers, and third-party payers in identifying health services for patients with LBP.
From the FULL TEXT Article:
Eighty percent of adults experience back pain at some time in their lives. [1, 2] Of those seeking professional care, 70% of patients will go to either a primary care medical physician (MD) or a chiropractor (DC).  In the case of chronic low back pain (LBP), utilization can be dramatic. In a North Carolina study of chronic LBP, 91% of care seekers saw an MD and 25% saw a DC.  Overall, an estimated 11% of Americans visit a chiropractor each year, predominantly for back pain. [5, 6]
While the value of randomized controlled trials (RCT) is unquestioned, they have not translated well into practice to support the decisions that family physicians must make in the care of patients who have undifferentiated illnesses embedded in the psychosocial context of their lives. [7, 8] Cause and effect conclusions carry the greatest weight when derived from a well-done RCT. But while RCTs carried out with randomization, controlled interventions, and rigid specifications on patient eligibility are characterized by high reliability, they are too often limited in generalizability. [9-12]
In contrast, practice-based research (PBR) has less reliability, but its strength is in its much broader generalizability. [8, 13, 14] There is a trade-off. The choice of methodology should be dependent on which of these 2 characteristics is more important given the research questions that are being asked. Practice-based studies are an appropriate methodology for questions involving normative data on practice patterns and long-term follow-up. [7, 15] Brook and Lohr  emphasize that there is a need to integrate the findings from studies of efficacy with the findings of research on effectiveness, process (practice activities), and outcomes of care. It is in the practice settings where patients, technology, and physicians interact that we can complete our understanding of the clinical spectrum of illness as it affects most people in our community most of the time. [7, 8, 15]
We therefore used a practice-based, observational design [17, 18] to assess a broad range of sociodemographic, psychosocial, health status, economic, and care-seeking variables. [19-26] This article describes pain and disability outcomes from 2 weeks to 4 years in 4 cohorts: acute and chronic DC and MD patients with LBP. Emphasis is placed on comparison of DC and MD patients, comparison of acute and chronic patients, and trends in the 4-year data. Predictors of outcomes are also discussed.
It is well known that nonrandomized comparison studies are susceptible to bias from imbalance in unknown factors that cannot be accounted for in the analysis. On the other hand, generalization of randomized trials and systematic reviews to clinical practice certainly runs analogous risks. It is therefore instructive to synthesize systematic reviews of randomized trials and large practice-based, observational studies to identify trends in patient outcomes. However, the following caveats must be considered. Systematic reviews of spinal manipulation have limited their scope to the efficacy of manipulation and do not address chiropractic care per se.  The vast majority of chiropractic patients in our study did receive manipulation, but half also received some form of physical therapy. [24-25] It should also be noted that our study addresses relative effectiveness only, whereas the systematic reviews draw conclusions from studies of both efficacy and relative efficacy. Our study is indicative of efficacy only in as much as it can compare favorably with standard, accepted medical care.
Eight systematic reviews of clinical trials have addressed the efficacy of spinal manipulation for the treatment of LBP. [61-68] Of these, 4 found inconclusive evidence for efficacy of manipulation for chronic LBP, although none found evidence of inefficacy or advantage for standard medical care. [61-65] Later reviews, however, found moderate to strong evidence that manipulation was better than placebo, general medical practice, massage, bed rest, and analgesics. [66-68] In addition, Vroomen et al  note some evidence for efficacy in the treatment of sciatica. Four systematic reviews favored manipulation for LBP in acute patients, [61-63, 66, 67] while one found the evidence inconclusive.  One favorable systematic review did not distinguish acute from chronic LBP.  Summarizing the current state of knowledge, Bronfort et al  note evidence for efficacy of manipulation for acute and chronic LBP, particularly for acute patients. They further conclude that the evidence for long-term efficacy is inconclusive.
Our study generally supports the findings of systematic reviews of spinal manipulation for a profession whose signature therapy is manipulation. Interestingly, an advantage for DC care was notable for chronic patients, a population for which systematic reviews have been more circumspect. The relative effect appears to be sustained in the long-term up to 12 months. The advantage for DC care over MD care for acute patients was small and not clinically important. This is not inconsistent with the findings of Carey et al.  in a previous large observational study that concluded comparability of care. Our study also suggests promising avenues for further research into the efficacy and relative effectiveness of DC care. Perhaps the greatest potential for DC care is in treatment of LBP with concomitant pain radiating below the knee. Chronic care for LBP in general should be explored.
The time trends in patient outcomes raise the question as to why a diminution of sustained pain relief is observed between 12 and 24 months Fig 1, Fig 3. This occurs despite the fact that almost half of the patients had sought care within the previous year at the 36-month and 48-month follow-ups. Further study is also merited into the enduring benefit of acute care over chronic care seen at 3 years in terms of pain intensity and days experienced, despite the fact that most patients, acute or chronic, have recurrent back problems. Continuous monitoring would shed light on the relationship of outcomes intensity, duration, and health care utilization. What is clear is that our findings support the contention that LBP persists for many patients far longer than previously believed. [56, 72-73] A large proportion of LBP patients improve under the care of chiropractors and medical doctors. However, it is apparent that advancement in patient management is required to broaden the effectiveness of both professions.
Our study supports the generalizability of systematic reviews of the efficacy of spinal manipulation for pain and functional disability to the effectiveness of chiropractic care in clinical practice. In terms of relative effectiveness, chiropractic care demonstrated advantage over medical care for chronic patients in the first year, particularly for those with leg pain radiating below the knee. Most of the relief was achieved within 3 months and sustained for 12 months; greater improvement was seen in acute patients. The exacerbation of LBP after 12 months remains to be explored. Studies of LBP do not generally capture the effect of care on the episodic nature of LBP. Future studies should include more continuous monitoring of outcomes. Predictors of outcome were consistent with those reported previously: baseline severity, duration of LBP at baseline, sciatica especially with pain below the knee, and physician type. Early intervention in an LBP episode is suggested by the relative benefit of any care for acute patients over chronic patients in terms of outcome intensity and pain days.
On average, patients with LBP patients improve under the care of either DC or MD care. Patient choice and interdisciplinary referral should be prime considerations by physicians, policymakers, and third-party payers in identifying health services for back patients.